This article overviews the process for configuring automatic failover of DNS endpoints in your network — meaning, if one DNS endpoint goes down, the NS1 platform will automatically redirect traffic to the next available endpoint.
Note
Refer to Configuring manual DNS failover for instructions of setting up manual failover (as opposed to automatic) which can be used for CNAME records or other record types compatible with a failover configuration.
Refer to any of the following articles for instructions on creating an NS1 monitor (i.e., probe):
Alternatively, you can configure a data source from NS1's API webhook or a third-party monitoring service integration. Refer to this section for integration-specific guides.
Note
The example used in this article leverages NS1's built-in monitors to configure automatic failover.
Note
Refer to Create a DNS record for detailed instructions and more information about all the DNS record types supported by NS1.
-
In the NS1 portal, navigate to DNS > Zones to view a list of zones associated with your account.
-
Click the name of the zone from the list to drill down into zone details, including a list of records if any.
-
Click Add Record.
Select a record type from the drop-down menu, and then enter the subdomain. (Note that you can leave the domain field blank to create a root-level record.) Optionally, adjust the time-to-live (TTL) value in seconds (default is 3600 seconds).
-
Enter one or more answers to the record. The requested information for each answer depends on the type of record you are creating. When creating a record for automatic failover, typically, each answer represents an endpoint to which to distribute traffic. These answers are filtered through the Filter Chain (once configured) to determine the optimal endpoint for each DNS query.
Note
For more information on the different types of records and their supported answers, refer to Create a DNS record.
Once complete, click Save All Changes. The new record appears in the list.
-
Click the name of the record you just created to drill down into record details, and then click Create Filter Chain.
-
Add the Up filter, Priority filter, and the Select First N filter in that order. Then, click Save Filter Chain.
The new Filter Chain processes answers by removing any answers that are "down" (i.e., up=false), then rearranging the answers based on their assigned priority value, and then reducing the list to the first N (number) of answers in the answer pool.
Next, you must adjust the answer metadata — connecting the monitor(s) or data source(s) you created in step 1 to their corresponding answers.
To complete configuration, you must edit the metadata for each answer to connect the corresponding monitor or data source for automatic updates as well as to assign a priority value.
-
Back on the record details page, click the Up filter to see the corresponding metadata label appear beneath each answer. Then, click the label beneath the first answer to display the answer metadata modal.
-
Click the "feed" icon and select the monitor or data feed to associate with this answer. Then, click Ok. Repeat this step for all remaining answers.
Note
If you have other DNS records serviced by the same servers, you can reuse the same data feeds in those records. Updates to a data feed are pushed automatically to all associated answers.
-
Next, back on the record details page, click the Priority filter to reveal the "priority" metadata field beneath the first answer and assign a priority value. A lower number indicates a higher preference when making traffic routing decisions. Repeat this step for all remaining answers.
Click Save Record.
Now, the monitor or data source you assigned to each answer will keep it up to date. Upon each DNS query, the answers are filtered through the Filter Chain using the following logic: First, all "down" answers (i.e., endpoints) are removed from the answer pool. Then, the answers are arranged based on their assigned priority value. Finally, all but the first "N" number of answers are removed from the list and the remaining answer(s) are returned to the requester. By default, N is set to 1 meaning only the first answer (i.e., available and highest priority) is returned to the requester.